Walking Patterns Reveal Alzheimer’s Risk Differently in Women and Men, Study Finds

Alzheimer’s disease has long been known to affect women and men differently, from the speed at which memory declines to the way the brain responds to the same underlying pathology. A new study published in BMC Medical Imaging adds a striking and unexpected dimension to this picture: in the earliest, pre-symptomatic stage of the disease, the same amyloid burden appears to reshape the brain and alter the way people walk in fundamentally different ways depending on sex. The findings come from a team of researchers based primarily at Nanjing Drum Tower Hospital in China, who set out to answer a deceptively simple question: when beta-amyloid, the protein widely considered a hallmark of Alzheimer’s disease, begins to accumulate in the brain, does it leave the same fingerprints on gait and brain structure in women as it does in men?

The answer, according to the new research, is a resounding no. In women with evidence of amyloid pathology, the team found measurable changes in walking speed and, remarkably, a negative relationship between the volume of a specific subregion of the hippocampus and several fine-grained gait parameters. In men with the same amyloid status, there were no gait changes at all, but there was a distinct pattern of atrophy in a different part of the hippocampus. The study suggests that the path from early Alzheimer’s pathology to observable physical changes may follow sex-specific routes through the brain, a conclusion with potentially significant implications for how clinicians screen for and monitor the prodromal phase of the disease.

Gait has attracted growing attention in dementia research over the past decade because it is increasingly understood as a window into brain function. Walking is not the automatic, mechanical process it might appear to be; it depends on coordinated activity across motor cortex, basal ganglia, cerebellum, and higher-order cognitive networks, including the hippocampus and the frontoparietal control systems that govern attention and executive function. Subtle disruptions in any of these circuits can show up as changes in stride length, cadence, swing velocity, or step-to-step variability long before a person notices anything wrong. Because gait can be measured cheaply, objectively, and repeatedly, it has become an appealing candidate for a non-invasive early indicator of neurodegeneration, particularly in the prodromal stage of Alzheimer’s disease, when amyloid plaques are accumulating but cognition remains normal or only mildly impaired.

To investigate whether sex modifies the relationship between amyloid pathology and gait, the researchers conducted a cross-sectional study of 253 participants aged 55 to 85 years, all of whom were either cognitively normal or had mild cognitive impairment. The cohort included 96 men and 157 women. Every participant underwent a comprehensive battery of assessments: quantitative gait testing, measurement of plasma beta-amyloid 42 and 40 concentrations, neuropsychological evaluation, and structural magnetic resonance imaging of the brain. The plasma Aβ42/40 ratio served as the key biomarker for amyloid pathology, allowing the team to classify participants as amyloid-positive or amyloid-negative without resorting to invasive lumbar punctures or expensive PET scans.

With participants stratified by both sex and amyloid status, the researchers used general linear models to compare gait parameters and gray matter structure across groups, applying false discovery rate correction to guard against spurious statistical findings. The results in women were the most eye-catching. Amyloid-positive females exhibited faster gait speeds than their amyloid-negative counterparts, an initially counterintuitive observation that the team linked to structural changes in the right hippocampal head. Specifically, in the amyloid-positive women, the volume of the right CA1 head, a subfield of the hippocampus critical for memory processing, showed a significant negative correlation with several right-sided gait measures: swing velocity average (r = −0.51, p = 0.003), minimum swing velocity (r = −0.36, p = 0.046), and minimum step speed (r = −0.36, p = 0.042). In plain terms, the larger the right hippocampal head, the slower and less dynamic the swing phase of the right leg.

No such correlations appeared in the amyloid-negative women, whose gait measures and neuroimaging parameters were statistically unrelated. This dissociation is important because it suggests the brain-gait coupling observed in the amyloid-positive women is not a general feature of aging or of female physiology, but something that emerges specifically in the presence of Alzheimer’s pathology. The authors propose that the pattern might be attributed to enlargement of the right hippocampal head, a finding that raises intriguing questions about whether early amyloid-related changes in this region could reflect compensatory processes, inflammation, or altered neural dynamics that manifest as faster but structurally coupled gait changes. Whatever the mechanism, the practical implication is that gait, combined with hippocampal subfield volumetry, could serve as a sex-informed marker of early amyloid-related brain change in women.

The story in men was different in almost every respect. Amyloid-positive males did not show altered gait, but they did exhibit significant atrophy of the left hippocampal body compared with amyloid-negative males. This left-sided, body-region atrophy contrasts sharply with the right-sided, head-region enlargement-associated pattern seen in women, underscoring that the same systemic pathology, detected by the same plasma biomarker, is associated with anatomically and functionally distinct consequences in each sex. The absence of gait changes in amyloid-positive men suggests that hippocampal body atrophy at this stage may not yet be severe enough, or may not involve the right circuitry, to disrupt the motor networks that govern walking.

Why might such pronounced sex differences exist? The authors and the broader literature point to several converging possibilities. Women have a higher lifetime risk of Alzheimer’s disease, in part because of longevity but also because of hormonal, genetic, and immune factors; the apolipoprotein E ε4 allele, the strongest genetic risk factor for late-onset disease, exerts a stronger effect in women. Sex hormones such as estrogen influence amyloid processing, tau phosphorylation, and hippocampal plasticity, and their decline after menopause may change how the female brain responds to accumulating pathology. There are also baseline sex differences in hippocampal volume, gait kinematics, and the organization of motor-cognitive networks, meaning that the same pathological insult lands on different neural substrates in men and women. The new study’s finding that amyloid-positive women show right hippocampal head changes while amyloid-positive men show left hippocampal body atrophy fits this framework of divergent vulnerability.

The technical approach of the study deserves attention as well. Rather than relying on a single crude measure such as overall walking speed, the team captured a rich panel of spatiotemporal gait parameters, including swing velocity averages and minima and step speed minima on each side of the body, and paired these with fine-grained hippocampal subfield volumetry rather than whole-hippocampus measures. This granularity matters because the hippocampus is not a uniform structure; its subfields, including the CA1 region of the head and the body, have distinct connectivity patterns and distinct patterns of vulnerability in Alzheimer’s disease. By correlating subfield-specific volumes with side-specific gait parameters, the study demonstrates the kind of precision that may be necessary to detect the earliest functional consequences of amyloid pathology, changes that would be invisible to coarser measures.

The study does have limitations inherent to its design. It is cross-sectional, meaning it captures a single moment in time and cannot establish whether the gait changes in amyloid-positive women precede, accompany, or follow the hippocampal changes, nor can it predict who will progress to symptomatic Alzheimer’s disease. The cohort, drawn from a Chinese clinical registry with ethics approval from the Alzheimer’s Disease Registry Platform and conducted under the Declaration of Helsinki, may not generalize to all populations. Still, the implications are compelling. If the effects of amyloid on gait are genuinely specific to women, as the authors conclude, then gait assessment could become a particularly valuable, low-cost screening tool for early Alzheimer’s disease in female patients, while men may require different markers, such as direct hippocampal body volumetry, to detect the same underlying pathology. As plasma biomarkers make amyloid testing increasingly accessible, studies like this one move the field closer to a personalized, sex-informed model of detecting Alzheimer’s disease decades before dementia takes hold.

Subject of Research: Sex-specific effects of beta-amyloid pathology on gait and hippocampal structure in prodromal Alzheimer's disease

Article Title: Sex-specific dynamics of gait and hippocampal structure in prodromal Alzheimer’s disease

Article References: Fang, S., Mo, Y., Pan, Y., Xie, P., Xiong, R., Mao, C., Jiang, J., Zhu, X., & Xu, Y. (2026). Sex-specific dynamics of gait and hippocampal structure in prodromal Alzheimer’s disease. BMC Medical Imaging. https://doi.org/10.1186/s12880-026-02523-9

Image Credits: AI Generated

DOI: 10.1186/s12880-026-02523-9

Keywords: Alzheimer's disease, gait, sex differences, beta-amyloid, hippocampus, prodromal Alzheimer's disease, MRI, plasma biomarkers, mild cognitive impairment, neurodegeneration, CA1 subfield, early detection